To add a couple more points:Going back to Metabolic's...

To add a couple more points:

Going back to Metabolic's announcement of 17 December 2004, management made a few statements that raised questions in my mind, and those questions have come back into sharp focus now.

For example, "Metabolic is confident and our FDA consultant has confirmed that the results already in hand have sufficient efficacy data to satisfy the requirements of the FDA to undertake the Phase 3, preceded by the "lead-in" dose confirmation period." Now, back then, we were told that this dose confirmation study would "add no more than a few months to the development path."

In point of fact, a more substantial and robust program was undertaken, but still it was always sold as a "dose-finding" study, not a re-consideration of the fundamental question of efficacy. Thus, I think I am justified in asking how such a fundamental error could have allowed to happen. Was the data coming from the 2004 trails in fact "sufficient efficacy data to satisfy the requirements of the FDA to undertake the Phase 3, or was it NOT?"

Or, put another way, how could it be possible that a "dose-finding" study of a late stage peptide drug, under any set of reasonable circumstances imaginable, result in a complete upset of ALL previous data gathering and conclusions drawn therefrom. Or, does it mean that previous data and the conclusions drawn therefrom were compromised or otherwise unreliable?

And if this is the case, why?

Here's another question that has nagged my mind: mention was made in the December 17th 2004 release that "There were several people in all AOD9604 treated groups who lost large amounts of weight (8kg or more and up to 21kgs) compared to none in placebo."

Now, that's great anecdotal information - but of questionable scientific value. The problem is that if such outlier values, like a 21 kg weight loss are included in the statistical analysis, it will significantly -and wrongly - influence the aggregate result.

We know that the 1mg dose group ended with 34 completers. If it so happens that the 21kg loser was in the 1mg group, and that data point was included in the calculation of the mean, it would have added a half kilo of average weight loss in this group and also helped improve the p value. This effect would be entirely spurious, as no one can seriously propose that they were measuring the effect of AOD9604 with this 21 kg loss. They actually measured the excessively zealous program of a diet fanatic, and that's all.

I want to know what dosage group this outlier belonged to, and whether this and other outlier values (whether high or low) were kept or discarded per best statistical practice.

I asked this very question of Metabolic in early 2005, and at that time they were unwilling to provide me the data, on the basis that it could not be made available pending publication of the data in a scientific peer-reviewed journal. After two years, I don't think that is any longer likely to happen, so I again ask, "Were outliers discarded from the data set in the 2004 study?"

If they were not, then everything that has happened for the past two years was based on a fundamental misapprehension as to the efficacy of this drug and from my perspective, it would look suspiciously intentional.

For me, the sad thing is that, looking again at all the data, from all the trial, even including the recent trial, it still looks to me like AOD9604 aid in the loss of weight. The latest trial shows an unusually long duration placebo effect due to the very effective diet and exercise regimen, perhaps masking the drug effect during most of the trial period.

You can see in the data that after week 16 treated groups continued to lose weight after the placebo group had stopped. What would another 12 weeks shown?

Testing a diet drug is a tricky business because there are so many social and behavioral variables. This drug may well have succeeded in a longer Phase 3 trial. But now, we'll almost certainly never know.

The only way we'll ever find out now is if the osteoporosis indication is commercialised, and we see weight loss as a side effect of the osteoporosis treatment.

That would be ironic indeed.